> ## Documentation Index
> Fetch the complete documentation index at: https://docs.llmeasy.ru/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect Hermes Agent to LLMEasy with a custom model

> Connect LLMEasy to Hermes Agent with GPT Key group models and the OpenAI-compatible protocol.

Hermes Agent is an AI agent for local terminal workflows. You can use it as a daily development assistant, or combine skills, MCP, scheduled tasks, and subagents for longer automation work.

When connecting LLMEasy, use a **GPT Key group** API Key and call models through the OpenAI-compatible endpoint.

<Info>
  This page only covers the GPT Key group setup for Hermes Agent. Copy model IDs from the <a href={"https://www.llmeasy.ru/pricing"}>model plaza</a>.
</Info>

## Prerequisites

* LLMEasy API Key (<a href={"https://www.llmeasy.ru/register"}>register here</a>)
* The API Key belongs to the **GPT Key group**
* A GPT Key group model ID copied from the <a href={"https://www.llmeasy.ru/pricing"}>model plaza</a>
* macOS, Linux, or Windows WSL2

## Install Hermes Agent

<Tabs>
  <Tab title="macOS / Linux / WSL2">
    ```bash theme={null}
    curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
    ```
  </Tab>

  <Tab title="Windows">
    On Windows, prefer WSL2 first. Native PowerShell support depends on your current Hermes Agent version. If you use native Windows, follow the installer prompts for your installed version.
  </Tab>
</Tabs>

## Setup

<Tabs>
  <Tab title="Interactive setup">
    <Steps>
      <Step title="Run the model selector">
        ```bash theme={null}
        hermes model
        ```
      </Step>

      <Step title="Choose the OpenAI or OpenAI Compatible provider">
        Select **OpenAI** or **OpenAI Compatible** from the provider list.
      </Step>

      <Step title="Fill in the LLMEasy config">
        Enter:

        * **Base URL**: `https://www.llmeasy.ru/v1`
        * **API Key**: your LLMEasy API Key
        * **Model**: a **GPT Key group** model ID copied from the <a href={"https://www.llmeasy.ru/pricing"}>model plaza</a>
      </Step>

      <Step title="Restart Hermes Agent">
        After updating the config, quit the current Hermes Agent session and start it again.
      </Step>

      <Step title="Start Hermes Agent">
        ```bash theme={null}
        hermes
        ```
      </Step>
    </Steps>
  </Tab>

  <Tab title="CLI config">
    If you prefer copy-paste commands, write the Hermes Agent config directly.

    ```bash theme={null}
    hermes config set llm.provider openai
    hermes config set llm.base_url "https://www.llmeasy.ru/v1"
    hermes config set llm.api_key "YOUR_LLMEASY_API_KEY"
    hermes config set llm.model "YOUR_GPT_MODEL_ID"
    ```

    Replace `YOUR_LLMEASY_API_KEY` with your LLMEasy API Key, and replace `YOUR_GPT_MODEL_ID` with a GPT Key group model ID from the <a href={"https://www.llmeasy.ru/pricing"}>model plaza</a>.

    Then start Hermes Agent:

    ```bash theme={null}
    hermes
    ```
  </Tab>

  <Tab title="Environment variables">
    If you do not want to store the API Key in the Hermes Agent config, export it in the current shell.

    ```bash theme={null}
    export OPENAI_API_BASE="https://www.llmeasy.ru/v1"
    export OPENAI_API_KEY="YOUR_LLMEASY_API_KEY"
    export HERMES_MODEL="YOUR_GPT_MODEL_ID"
    hermes
    ```

    These variables only apply to the current shell session. Export them again in a new terminal, or add them to your own shell profile.
  </Tab>
</Tabs>

## Feature quick reference

| Feature         | What it is for                                 |
| --------------- | ---------------------------------------------- |
| Terminal UI     | Chat with the agent directly in your terminal  |
| Skills          | Save reusable workflows for repeated tasks     |
| MCP             | Connect local or remote tool capabilities      |
| Scheduled tasks | Run timed jobs and recurring workflows         |
| Subagents       | Split complex work across focused agents       |
| Message gateway | Connect the agent to more message entry points |

## Migrate from OpenClaw

If you already use OpenClaw, first check whether your current Hermes Agent version provides the migration command. Run a dry run before changing anything.

```bash theme={null}
hermes claw migrate --dry-run
```

After you confirm the output, run the migration:

```bash theme={null}
hermes claw migrate
```

After migration, still review the Hermes Agent **Base URL**, API Key, and model ID. They should match `https://www.llmeasy.ru/v1`, your LLMEasy API Key, and a GPT Key group model ID.

## Troubleshooting

* If requests fail or return no response, confirm that **Base URL** is exactly `https://www.llmeasy.ru/v1`. Do not append `/chat/completions` in this field.
* If Hermes reports that the model does not exist, confirm that **Model** is a GPT Key group model ID from the <a href={"https://www.llmeasy.ru/pricing"}>model plaza</a>.
* If authentication fails, check for copied spaces and confirm that the API Key belongs to the **GPT Key group**.
* If Hermes still uses the old model after a change, quit the current Hermes Agent session and run `hermes` again.
* LLMEasy usually does not require an extra `User-Agent` or custom headers. Only add headers when your custom gateway, tenant routing, observability, or caching layer explicitly requires them.

<Note>
  Do not commit real API Keys to your repository. For team-shared setup, prefer environment variables or local config.
</Note>

## Related FAQ

* [MCP vs API Key and Base URL](/en/faq/concepts/mcp-vs-api-key-base-url)
* [What are model\_provider, base\_url, and wire\_api?](/en/faq/codex/model-provider-base-url-wire-api)
* [How to configure an OpenAI-compatible API in Cline](/en/faq/cline/openai-compatible-api)
* [OpenClaw LLMEasy setup guide](/en/ai-tools/openclaw)